SPINE Volume 40, Number 8, pp 570-579 ©2015, Wolters Kluwer Health, Inc. All rights reserved.
LITERATURE REVIEW
Comparison of Smith-Petersen Osteotomy and Pedicle Subtraction Osteotomy for the Correction of Thoracolumbar Kyphotic Deformity in Ankylosing Spondylitis A Systematic Review and Meta-analysis Hui Liu, MD, Changsheng Yang, MD, Zhaomin Zheng, MD, PhD, Wenbin Ding, MD, Jianru Wang, MD, Hua Wang, MD, and Sibei Li, MD
Study Design. A systematic review and meta-analysis. Objective. To compare the efficacy and safety outcomes of SmithPetersen osteotomy (SPO) and pedicle subtraction osteotomy (PSO) for patients with ankylosing spondylitis (AS) with thoracolumbar kyphotic deformity. Summary of Background Data. Both SPO and PSO are used to correct thoracolumbar kyphotic deformity due to AS. Evidence is insufficient to determine which has better efficacy and safety outcomes. Methods. A systematic literature search was performed. Studies concentrating on treating thoracolumbar kyphotic deformity due to AS with SPO and/or PSO were included. Efficacy was determined with radiographical outcomes, including sagittal vertical axis and lumbar lordosis. Safety was determined with complication rates. The data were analyzed with Review Manager and R software. Results. Twenty-three studies were included. Among them, 5 were comparative studies and were used for a meta-analysis. All 23 studies were pooled to evaluate the radiographical correction and incidence of complications. The meta-analysis of the 5 comparative studies showed no significant difference between groups in either correction of sagittal vertical axis and lumbar lordosis or incidence of complications. The pooled data also showed similar radiological correction and complication rates between SPO and PSO. SPO was found to have potential risk of aortic rupture and slightly higher risk From the Department of Spine Surgery, The First Affiliated Hospital, Sun Yat-sen University, Guangzhou, China. Acknowledgment date: October 12, 2014. Revision date: January 15, 2015. Acceptance date: January 28, 2015. The manuscript submitted does not contain information about medical device(s)/drug(s). No funds were received in support of this work. No relevant financial activities outside the submitted work. Address correspondence and reprint requests to Zhaomin Zheng, MD, PhD, Department of Spine Surgery, The First Affiliated Hospital, Sun Yat-sen University, 58 Zhongshan 2nd Rd, Guangzhou, 510080, China; E-mail: [email protected] DOI: 10.1097/BRS.0000000000000815
570
of permanent neurological deficit without statistical significance, whereas PSO was shown longer operative time and more blood loss. Conclusion. This systematic review and meta-analysis demonstrates that both SPO and PSO are effective in correcting thoracolumbar kyphotic deformity in AS and have similar risk of most complications. Aortic rupture and related death during correction is reported in SPO and should be taken into consideration for decision making. Key words: Smith-Petersen osteotomy, pedicle subtraction osteotomy, ankylosing spondylitis, systematic review, meta-analysis. Level of Evidence: 4 Spine 2015;40:570–579
A
nkylosing spondylitis (AS) is a chronic immunemediated disease primarily affecting the axial skeleton. The prevalence of AS is between 0.1% and 0.4%, and there is a male predominance.1 The etiology of AS remains unclear, but a close correlation between human leukocyte antigen B27 and AS has been observed.2 Thoracolumbar kyphotic deformity is a disabling condition affecting more than 30% of patients with AS.3 Structurally, as the disease progresses, the lumbar lordosis flattens, with the head and neck thrusting forward. The patients find it difficult to lie flat, stand upright, and look straight forward due to this sagittal imbalance. In severe cases, the spinal deformity may even lead to compression of the abdomen and restricted respiratory function. Surgical intervention is usually required in the advanced stage of AS to restore the sagittal balance to achieve erect posture and the ability to see straight ahead. Osteotomy is extensively used to correct thoracolumbar kyphotic deformity and improve the spinal-pelvic sagittal alignment. Different types of osteotomies, such as SmithPetersen osteotomy (SPO),4 pedicle subtraction osteotomy (PSO),5 polysegmental wedge osteotomy,6 closing-opening wedge osteotomy,7 and transpedicular bivertebrae wedge osteotomy and discectomy,8 have been reported. SPO and PSO are the most common choices in the literature.
www.spinejournal.com
April 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141337_LR 570
16/03/15 8:25 PM
LITERATURE REVIEW
Comparison of Smith-Petersen Osteotomy and Pedicle Subtraction Osteotomy • Liu et al
SPO was first described in 1945 by Smith and Petersen as an osteotomy for the treatment of kyphotic deformity.4 In 1968, Goel9 performed SPO in patients with AS and achieved a satisfying correction. Since then, 1-level SPO has been continually reported in the treatment of AS.10–19 In SPO, all posterior ligaments and facets are dissected, and the posterior column is shortened whereas the anterior column is lengthened by opening the ossified disc space to achieve lumbar lordosis. PSO is a more popular choice in recent years. A vertebral wedge osteotomy for the correction of thoracolumbar kyphotic deformity in patients with AS was first described by Scudese in 1963.5 In PSO, both of the pedicles and a V-shaped bony wedge of the vertebral body are resected. The posterior column is shortened without lengthening the anterior column. According to the classification of osteotomies, PSO is technically more demanding and is a higher-grade osteotomy than SPO.20 Many publications have shown that both techniques provide satisfactory correction outcomes and have thoroughly reported the technical notes, advantages, disadvantages, indications, and contraindications of these 2 types of osteotomies. However, few studies have directly compared the 2 techniques, and the results were controversial. Chang et al21 and Zhu et al22 observed no significant difference in sagittal vertical axis (SVA) or lumbar lordosis (LL) between SPO and PSO, whereas Lazennec et al10 and Willems et al11 reported a greater correction of LL in PSO. The reported complication rates varied greatly, making it difficult to estimate the safety outcomes of the 2 osteotomies. Furthermore, except for Van Royen’s well-known review, which found no solid information to aid in decision making,23 no studies to our best knowledge have provided a systematic review of osteotomies in patients with AS with thoracolumbar kyphotic deformity. The purpose of the current study is to conduct a systematic review and meta-analysis to evaluate the efficacy and safety outcomes of SPO and PSO for correction of thoracolumbar kyphotic deformity in AS, with the aim of providing more information to aid in clinical decision making.
MATERIALS AND METHODS A systematic review and meta-analysis in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines was performed with a predetermined protocol.
Inclusion and Exclusion The inclusion criteria were (1) patients with thoracolumbar kyphotic deformity attributable to AS treated by SPO or PSO at the lumbar or lower thoracic segments; (2) descriptive data such as age, level, and number of surgical procedures available; and (3) radiographical outcomes (at least 5 patients) and/or data related to complication available (at least 10 patients). Studies were excluded according to the following criteria: (1) duplicate publication; (2) patients were treated with more than 1 type of osteotomy; (3) patients with AS had osteotomy due to another etiology, such as spinal tuberculosis Spine
or post-traumatic kyphosis; and (4) patients had osteotomy at the cervical or the upper thoracic segments.
Literature Search With the terms “osteotomy” and “ankylosing spondylitis,” we searched PubMed, EMBASE, and the Cochrane Database for all articles written in English from inception to November 1, 2013. The reference lists of related reviews were also screened for potentially missing studies.
Study Selection and Data Extraction Two authors independently confirmed the eligibility of studies and extracted the data. Any divergences were resolved by a third author.
Assessment of Study Quality The quality of the included studies was assessed with the MINORS tool.24 There were 8 items, and an additional 4 items were added for comparative studies. The global ideal score was 16 for noncomparative studies and 24 for comparative studies. Two authors independently assessed study quality, and divergences were resolved by a third author.
Statistical Analysis The data from comparative studies were primarily analyzed. Radiographical outcomes and complication rates were directly compared. The data from noncomparative studies together with comparative studies were pooled to evaluate the radiographical correction and incidence of complications. Continuous data were pooled with the weighted mean difference and dichotomous data with odds ratios (ORs). All analyses were performed with the aid of Review Manager software from the Cochrane Collaboration (Version 5.2) or R software (Version 3.0.2). The results are reported as the mean with 95% confidence interval (CI). Statistical heterogeneity among studies was evaluated using the I2 statistic, judging values of less than 25% to be minimal, 25% to 50% to be moderate, and 50% or greater to be substantial. If I2 was judged to be minimal or moderate, a fixed-effects model was employed. Otherwise, a random-effects model was suitable. All tests were 2-tailed, and a P value of less than 0.05 was deemed statistically significant.
RESULTS The search strategy identified 228 publications. Titles and abstracts were screened for inclusion. Fifty-four were retrieved for more detailed assessment, from which 31 were excluded for the following reasons: SPO for pseudoarthrosis (2), osteotomy at the level of C7 (1), closing-opening wedge osteotomy (1), duplicate publication (2), data not extractable (10), insufficient number of patients (14), and patients had both SPO and PSO (1) (Figure 1). Twenty-three studies were included: 5 comparative studies,10,11,21,22,25 7 SPO studies,9,12–19 and 11 PSO studies.26–36 No randomized controlled trials were included. All studies except 2 (1 SPO and 1 PSO) were retrospective. www.spinejournal.com
571
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141337_LR 571
16/03/15 8:25 PM
LITERATURE REVIEW
Comparison of Smith-Petersen Osteotomy and Pedicle Subtraction Osteotomy • Liu et al
(I2 = 39%, mean difference = −0.06°, P = 0.98) or loss of LL (I2= 0%, mean difference = 0.95°, P = 0.12) (Figure 4). Comparison of Complications According to 5 studies, there was no significant difference between the groups in total complications (I2 = 0%, OR = 1.55, 95% CI: 0.92–2.62, P = 0.10), neurological deficit (I2 = 0%, OR = 0.77, 95% CI: 0.33–1.80, P = 0.54), or dural tear (I2 = 0%, OR = 1.82, 95% CI: 0.84–3.95, P = 0.13). Only 2 studies reported implant-related complications. No significant difference was observed between the groups (I2 = 0%, OR = 1.18, 95% CI: 0.44–3.19, P = 0.74) (Figure 5). Figure 1. Flow diagram of study selection. SPO indicates SmithPetersen osteotomy; AS, ankylosing spondylitis; PSO, pedicle subtraction osteotomy.
Table 1 shows descriptive data for the 23 qualifying studies. A total of 979 patients (441 for SPO and 538 for PSO) were included. The average age varied from 32.6 to 54.7 years. All osteotomies were performed in the lower thoracic and lumbar segments (T11–L5). Most osteotomies were performed at only 1 level. The indication for surgery was clearly stated as spinal deformity, such as inability to look straight ahead, walking difficulty, abdominal compression, and restricted respiratory function, in all of the studies.
Assessment of Study Quality
The study quality score was 14.8 ± 2.8 for comparative studies, 8.4 ± 1.7 for SPO studies, and 9.3 ± 1.8 for PSO studies. Overall, the included studies were of median quality.
Meta-analysis With the 5 Comparative Studies The 5 comparative studies were used for a meta-analysis to directly compare SPO and PSO (Figure 2). Three studies reported both the preoperative and follow-up SVA and LL, 2 of which also reported data of postoperative correction (correction immediately after operation) and loss of correction. All 5 studies analyzed the rates of complications. Comparison of SVA The preoperative SVA ranged from 13.6 cm to 14.7 cm in SPO and 14.5 cm to 18.5 cm in PSO. There was no significant difference between the groups in preoperative SVA (I2 = 47%, mean difference = −1.67 cm, P = 0.12) or follow-up SVA (I 2= 60%, mean difference = −1.79 cm, P = 0.06). There was no significant difference between the groups in postoperative correction of SVA (I2 = 0%, mean difference = 0.62 cm, P = 0.36) or loss of SVA (I2 = 15%, mean difference = 0.34 cm, P = 0.07) (Figure 3). Comparison of LL The preoperative LL ranged from −3° to 12.6° in SPO and −5° to 15.4° in PSO. There was no significant difference between the groups in preoperative (I2 = 43%, mean difference = −0.74°, P = 0.69) or final follow-up LL (I2 = 93%, mean difference = −7.54°, P = 0.36). There was no significant difference between the groups in postoperative correction of LL 572
Pooled Data From All 23 Studies As a supplement and to provide more detailed information, data from all 23 studies were pooled to evaluate the incidence and correction of complications from SPO and PSO (Figure 2). Correction of SVAs In SPO, the preoperative SVA ranged from 13.6 cm to 14.7 cm. The postoperative correction was 10.68 cm (95% CI: 9.27–12.09). The follow-up correction was 9.80 cm (95% CI: 8.17–11.72). In PSO, the preoperative SVA ranged from 9.4 cm to 27.3 cm. The postoperative correction was 12.36 cm (95% CI: 8.59–16.13). The follow-up correction was 11.76 cm (95% CI: 7.34–16.18). The 95% CIs of the postoperative and follow-up corrections overlapped (Table 2). Correction of LL In SPO, the preoperative LL ranged from −1° to 12.6°. The postoperative correction was 32.44° (95% CI: 17.75–47.13). The follow-up correction was 34.25° (95% CI: 27.06–38.53). In PSO, the preoperative LL ranged from −5° to 21.36°. The postoperative correction was 35.33° (95% CI: 32.69–37.98). The follow-up correction was 36.05° (95% CI: 32.56–39.54). The 95% CIs of the postoperative and follow-up correction overlapped (Table 2). Incidence of Complications Three PSO studies were excluded because of an insufficient number of patients (50°, persistent painful spondylodiscitis despite less degree of deformity
Severe disabling kyphosis with forward bending
Indication for Surgery
Inability to look ahead, walking difficulty
Flexion deformity
Inability to look ahead
Inability to look ahead
Inability to look ahead
Disturbance of balance, inability to walk, and stand upright.
Forward-bending deformity
Indication for Surgery
Inability to look ahead, walking difficulty
Inability to look ahead, SVA >8cm
A disturbed sagittal alignment
Inability to look ahead, walking difficulty
Inability to look ahead, walking difficulty
Indication for Surgery
LITERATURE REVIEW Comparison of Smith-Petersen Osteotomy and Pedicle Subtraction Osteotomy • Liu et al
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
573
16/03/15 8:25 PM
574
Pigge et al
29
n indicates number of patients; Age, mean and range of age of patients; M:F, ratio of male to female; pre-SVA, preoperative sagittal vertical axis; pre-LL, mean and standard deviation of preoperative lumbar lordosis; Level, the level of osteotomy operated; NO, the number of osteotomy operative on the patient; Fu, mean and range of follow-up years; SPO, Smith-Petersen osteotomy; PSO, pedicle subtraction osteotomy; GK, global kyphosis.
Inability to look ahead 2 9 Zhang et al35
41 (35–51)
8:1
27.3 ± 4.4
0.9 ± 6.9
T12–L3
3.3 (2–5.7)
Disturbed sagittal alignment, inability to look ahead 1 33 Qian et al34
36 (21–50)
30:3
…
19.4 ± 5.2
L1–L2
…
60° < GK
LITERATURE REVIEW
Comparison of Smith-Petersen Osteotomy and Pedicle Subtraction Osteotomy for the Correction of Thoracolumbar Kyphotic Deformity in Ankylosing Spondylitis A Systematic Review and Meta-analysis Hui Liu, MD, Changsheng Yang, MD, Zhaomin Zheng, MD, PhD, Wenbin Ding, MD, Jianru Wang, MD, Hua Wang, MD, and Sibei Li, MD
Study Design. A systematic review and meta-analysis. Objective. To compare the efficacy and safety outcomes of SmithPetersen osteotomy (SPO) and pedicle subtraction osteotomy (PSO) for patients with ankylosing spondylitis (AS) with thoracolumbar kyphotic deformity. Summary of Background Data. Both SPO and PSO are used to correct thoracolumbar kyphotic deformity due to AS. Evidence is insufficient to determine which has better efficacy and safety outcomes. Methods. A systematic literature search was performed. Studies concentrating on treating thoracolumbar kyphotic deformity due to AS with SPO and/or PSO were included. Efficacy was determined with radiographical outcomes, including sagittal vertical axis and lumbar lordosis. Safety was determined with complication rates. The data were analyzed with Review Manager and R software. Results. Twenty-three studies were included. Among them, 5 were comparative studies and were used for a meta-analysis. All 23 studies were pooled to evaluate the radiographical correction and incidence of complications. The meta-analysis of the 5 comparative studies showed no significant difference between groups in either correction of sagittal vertical axis and lumbar lordosis or incidence of complications. The pooled data also showed similar radiological correction and complication rates between SPO and PSO. SPO was found to have potential risk of aortic rupture and slightly higher risk From the Department of Spine Surgery, The First Affiliated Hospital, Sun Yat-sen University, Guangzhou, China. Acknowledgment date: October 12, 2014. Revision date: January 15, 2015. Acceptance date: January 28, 2015. The manuscript submitted does not contain information about medical device(s)/drug(s). No funds were received in support of this work. No relevant financial activities outside the submitted work. Address correspondence and reprint requests to Zhaomin Zheng, MD, PhD, Department of Spine Surgery, The First Affiliated Hospital, Sun Yat-sen University, 58 Zhongshan 2nd Rd, Guangzhou, 510080, China; E-mail: [email protected] DOI: 10.1097/BRS.0000000000000815
570
of permanent neurological deficit without statistical significance, whereas PSO was shown longer operative time and more blood loss. Conclusion. This systematic review and meta-analysis demonstrates that both SPO and PSO are effective in correcting thoracolumbar kyphotic deformity in AS and have similar risk of most complications. Aortic rupture and related death during correction is reported in SPO and should be taken into consideration for decision making. Key words: Smith-Petersen osteotomy, pedicle subtraction osteotomy, ankylosing spondylitis, systematic review, meta-analysis. Level of Evidence: 4 Spine 2015;40:570–579
A
nkylosing spondylitis (AS) is a chronic immunemediated disease primarily affecting the axial skeleton. The prevalence of AS is between 0.1% and 0.4%, and there is a male predominance.1 The etiology of AS remains unclear, but a close correlation between human leukocyte antigen B27 and AS has been observed.2 Thoracolumbar kyphotic deformity is a disabling condition affecting more than 30% of patients with AS.3 Structurally, as the disease progresses, the lumbar lordosis flattens, with the head and neck thrusting forward. The patients find it difficult to lie flat, stand upright, and look straight forward due to this sagittal imbalance. In severe cases, the spinal deformity may even lead to compression of the abdomen and restricted respiratory function. Surgical intervention is usually required in the advanced stage of AS to restore the sagittal balance to achieve erect posture and the ability to see straight ahead. Osteotomy is extensively used to correct thoracolumbar kyphotic deformity and improve the spinal-pelvic sagittal alignment. Different types of osteotomies, such as SmithPetersen osteotomy (SPO),4 pedicle subtraction osteotomy (PSO),5 polysegmental wedge osteotomy,6 closing-opening wedge osteotomy,7 and transpedicular bivertebrae wedge osteotomy and discectomy,8 have been reported. SPO and PSO are the most common choices in the literature.
www.spinejournal.com
April 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141337_LR 570
16/03/15 8:25 PM
LITERATURE REVIEW
Comparison of Smith-Petersen Osteotomy and Pedicle Subtraction Osteotomy • Liu et al
SPO was first described in 1945 by Smith and Petersen as an osteotomy for the treatment of kyphotic deformity.4 In 1968, Goel9 performed SPO in patients with AS and achieved a satisfying correction. Since then, 1-level SPO has been continually reported in the treatment of AS.10–19 In SPO, all posterior ligaments and facets are dissected, and the posterior column is shortened whereas the anterior column is lengthened by opening the ossified disc space to achieve lumbar lordosis. PSO is a more popular choice in recent years. A vertebral wedge osteotomy for the correction of thoracolumbar kyphotic deformity in patients with AS was first described by Scudese in 1963.5 In PSO, both of the pedicles and a V-shaped bony wedge of the vertebral body are resected. The posterior column is shortened without lengthening the anterior column. According to the classification of osteotomies, PSO is technically more demanding and is a higher-grade osteotomy than SPO.20 Many publications have shown that both techniques provide satisfactory correction outcomes and have thoroughly reported the technical notes, advantages, disadvantages, indications, and contraindications of these 2 types of osteotomies. However, few studies have directly compared the 2 techniques, and the results were controversial. Chang et al21 and Zhu et al22 observed no significant difference in sagittal vertical axis (SVA) or lumbar lordosis (LL) between SPO and PSO, whereas Lazennec et al10 and Willems et al11 reported a greater correction of LL in PSO. The reported complication rates varied greatly, making it difficult to estimate the safety outcomes of the 2 osteotomies. Furthermore, except for Van Royen’s well-known review, which found no solid information to aid in decision making,23 no studies to our best knowledge have provided a systematic review of osteotomies in patients with AS with thoracolumbar kyphotic deformity. The purpose of the current study is to conduct a systematic review and meta-analysis to evaluate the efficacy and safety outcomes of SPO and PSO for correction of thoracolumbar kyphotic deformity in AS, with the aim of providing more information to aid in clinical decision making.
MATERIALS AND METHODS A systematic review and meta-analysis in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines was performed with a predetermined protocol.
Inclusion and Exclusion The inclusion criteria were (1) patients with thoracolumbar kyphotic deformity attributable to AS treated by SPO or PSO at the lumbar or lower thoracic segments; (2) descriptive data such as age, level, and number of surgical procedures available; and (3) radiographical outcomes (at least 5 patients) and/or data related to complication available (at least 10 patients). Studies were excluded according to the following criteria: (1) duplicate publication; (2) patients were treated with more than 1 type of osteotomy; (3) patients with AS had osteotomy due to another etiology, such as spinal tuberculosis Spine
or post-traumatic kyphosis; and (4) patients had osteotomy at the cervical or the upper thoracic segments.
Literature Search With the terms “osteotomy” and “ankylosing spondylitis,” we searched PubMed, EMBASE, and the Cochrane Database for all articles written in English from inception to November 1, 2013. The reference lists of related reviews were also screened for potentially missing studies.
Study Selection and Data Extraction Two authors independently confirmed the eligibility of studies and extracted the data. Any divergences were resolved by a third author.
Assessment of Study Quality The quality of the included studies was assessed with the MINORS tool.24 There were 8 items, and an additional 4 items were added for comparative studies. The global ideal score was 16 for noncomparative studies and 24 for comparative studies. Two authors independently assessed study quality, and divergences were resolved by a third author.
Statistical Analysis The data from comparative studies were primarily analyzed. Radiographical outcomes and complication rates were directly compared. The data from noncomparative studies together with comparative studies were pooled to evaluate the radiographical correction and incidence of complications. Continuous data were pooled with the weighted mean difference and dichotomous data with odds ratios (ORs). All analyses were performed with the aid of Review Manager software from the Cochrane Collaboration (Version 5.2) or R software (Version 3.0.2). The results are reported as the mean with 95% confidence interval (CI). Statistical heterogeneity among studies was evaluated using the I2 statistic, judging values of less than 25% to be minimal, 25% to 50% to be moderate, and 50% or greater to be substantial. If I2 was judged to be minimal or moderate, a fixed-effects model was employed. Otherwise, a random-effects model was suitable. All tests were 2-tailed, and a P value of less than 0.05 was deemed statistically significant.
RESULTS The search strategy identified 228 publications. Titles and abstracts were screened for inclusion. Fifty-four were retrieved for more detailed assessment, from which 31 were excluded for the following reasons: SPO for pseudoarthrosis (2), osteotomy at the level of C7 (1), closing-opening wedge osteotomy (1), duplicate publication (2), data not extractable (10), insufficient number of patients (14), and patients had both SPO and PSO (1) (Figure 1). Twenty-three studies were included: 5 comparative studies,10,11,21,22,25 7 SPO studies,9,12–19 and 11 PSO studies.26–36 No randomized controlled trials were included. All studies except 2 (1 SPO and 1 PSO) were retrospective. www.spinejournal.com
571
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141337_LR 571
16/03/15 8:25 PM
LITERATURE REVIEW
Comparison of Smith-Petersen Osteotomy and Pedicle Subtraction Osteotomy • Liu et al
(I2 = 39%, mean difference = −0.06°, P = 0.98) or loss of LL (I2= 0%, mean difference = 0.95°, P = 0.12) (Figure 4). Comparison of Complications According to 5 studies, there was no significant difference between the groups in total complications (I2 = 0%, OR = 1.55, 95% CI: 0.92–2.62, P = 0.10), neurological deficit (I2 = 0%, OR = 0.77, 95% CI: 0.33–1.80, P = 0.54), or dural tear (I2 = 0%, OR = 1.82, 95% CI: 0.84–3.95, P = 0.13). Only 2 studies reported implant-related complications. No significant difference was observed between the groups (I2 = 0%, OR = 1.18, 95% CI: 0.44–3.19, P = 0.74) (Figure 5). Figure 1. Flow diagram of study selection. SPO indicates SmithPetersen osteotomy; AS, ankylosing spondylitis; PSO, pedicle subtraction osteotomy.
Table 1 shows descriptive data for the 23 qualifying studies. A total of 979 patients (441 for SPO and 538 for PSO) were included. The average age varied from 32.6 to 54.7 years. All osteotomies were performed in the lower thoracic and lumbar segments (T11–L5). Most osteotomies were performed at only 1 level. The indication for surgery was clearly stated as spinal deformity, such as inability to look straight ahead, walking difficulty, abdominal compression, and restricted respiratory function, in all of the studies.
Assessment of Study Quality
The study quality score was 14.8 ± 2.8 for comparative studies, 8.4 ± 1.7 for SPO studies, and 9.3 ± 1.8 for PSO studies. Overall, the included studies were of median quality.
Meta-analysis With the 5 Comparative Studies The 5 comparative studies were used for a meta-analysis to directly compare SPO and PSO (Figure 2). Three studies reported both the preoperative and follow-up SVA and LL, 2 of which also reported data of postoperative correction (correction immediately after operation) and loss of correction. All 5 studies analyzed the rates of complications. Comparison of SVA The preoperative SVA ranged from 13.6 cm to 14.7 cm in SPO and 14.5 cm to 18.5 cm in PSO. There was no significant difference between the groups in preoperative SVA (I2 = 47%, mean difference = −1.67 cm, P = 0.12) or follow-up SVA (I 2= 60%, mean difference = −1.79 cm, P = 0.06). There was no significant difference between the groups in postoperative correction of SVA (I2 = 0%, mean difference = 0.62 cm, P = 0.36) or loss of SVA (I2 = 15%, mean difference = 0.34 cm, P = 0.07) (Figure 3). Comparison of LL The preoperative LL ranged from −3° to 12.6° in SPO and −5° to 15.4° in PSO. There was no significant difference between the groups in preoperative (I2 = 43%, mean difference = −0.74°, P = 0.69) or final follow-up LL (I2 = 93%, mean difference = −7.54°, P = 0.36). There was no significant difference between the groups in postoperative correction of LL 572
Pooled Data From All 23 Studies As a supplement and to provide more detailed information, data from all 23 studies were pooled to evaluate the incidence and correction of complications from SPO and PSO (Figure 2). Correction of SVAs In SPO, the preoperative SVA ranged from 13.6 cm to 14.7 cm. The postoperative correction was 10.68 cm (95% CI: 9.27–12.09). The follow-up correction was 9.80 cm (95% CI: 8.17–11.72). In PSO, the preoperative SVA ranged from 9.4 cm to 27.3 cm. The postoperative correction was 12.36 cm (95% CI: 8.59–16.13). The follow-up correction was 11.76 cm (95% CI: 7.34–16.18). The 95% CIs of the postoperative and follow-up corrections overlapped (Table 2). Correction of LL In SPO, the preoperative LL ranged from −1° to 12.6°. The postoperative correction was 32.44° (95% CI: 17.75–47.13). The follow-up correction was 34.25° (95% CI: 27.06–38.53). In PSO, the preoperative LL ranged from −5° to 21.36°. The postoperative correction was 35.33° (95% CI: 32.69–37.98). The follow-up correction was 36.05° (95% CI: 32.56–39.54). The 95% CIs of the postoperative and follow-up correction overlapped (Table 2). Incidence of Complications Three PSO studies were excluded because of an insufficient number of patients (50°, persistent painful spondylodiscitis despite less degree of deformity
Severe disabling kyphosis with forward bending
Indication for Surgery
Inability to look ahead, walking difficulty
Flexion deformity
Inability to look ahead
Inability to look ahead
Inability to look ahead
Disturbance of balance, inability to walk, and stand upright.
Forward-bending deformity
Indication for Surgery
Inability to look ahead, walking difficulty
Inability to look ahead, SVA >8cm
A disturbed sagittal alignment
Inability to look ahead, walking difficulty
Inability to look ahead, walking difficulty
Indication for Surgery
LITERATURE REVIEW Comparison of Smith-Petersen Osteotomy and Pedicle Subtraction Osteotomy • Liu et al
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
573
16/03/15 8:25 PM
574
Pigge et al
29
n indicates number of patients; Age, mean and range of age of patients; M:F, ratio of male to female; pre-SVA, preoperative sagittal vertical axis; pre-LL, mean and standard deviation of preoperative lumbar lordosis; Level, the level of osteotomy operated; NO, the number of osteotomy operative on the patient; Fu, mean and range of follow-up years; SPO, Smith-Petersen osteotomy; PSO, pedicle subtraction osteotomy; GK, global kyphosis.
Inability to look ahead 2 9 Zhang et al35
41 (35–51)
8:1
27.3 ± 4.4
0.9 ± 6.9
T12–L3
3.3 (2–5.7)
Disturbed sagittal alignment, inability to look ahead 1 33 Qian et al34
36 (21–50)
30:3
…
19.4 ± 5.2
L1–L2
…
60° < GK
